Introduction

Patients undergoing lung resections can experience persistent postoperative pain, which is a known risk factor for chronic opioid use [1–4]. In the setting of the US opioid epidemic, which claimed over 80,000 lives via overdose in 2023, this is an incredibly relevant issue when considering patients for surgery and managing their postoperative pain [5,6]. After any surgical procedure, there are concerns about the risks of chronic opioid use, addiction, and an impact on patient quality of life. This is especially true within thoracic surgery. Thoracic surgery patients experience some of the highest rates of persistent post-operative new opioid usage compared to other surgical specialties, with early postoperative pain intensity being found to be predictive of chronic postoperative pain in this population [7]. Approximately 1 in 4 thoracic surgery patients will experience chronic post-surgical pain after a minimally invasive lung resection (MILR), emphasizing the pervasive impact of this issue within the specialty [8]. Current efforts to mitigate the detrimental effects of peri-operative opioid usage in thoracic surgery are limited by a lack of comprehensive information regarding the predictors of chronic pain [2,8–10].

Prior efforts to identify risk factors for chronic pain in thoracic surgery have been conducted with an emphasis on post-thoracotomy pain syndrome (PTPS) [11,12]. Nevertheless, as surgical practice for lung resections shifts predominantly to minimally invasive approaches, it becomes relevant to identifying risk factors associated with new persistent opioid use in this patient population [13–16]. Factors identified with an increased likelihood of chronic use after surgery include younger age, male sex, a history of substance use disorder, and cancer surgery [7]. Given the increased prevalence of video-assisted thoracoscopic surgeries (VATS) and robotic-assisted thoracoscopic surgeries (RATS), it is crucial to understand these factors in the context of MILR.

MILRs specifically have unique considerations regarding postoperative pain. There is a known neuropathic component in a subset of patients with postoperative pain, which is attributed to intercostal nerve injury or irritation [7,17,18]. Intercostal neuropraxia is specific to thoracic surgery and has been found to cause greater analgesia use and impact patients’ quality of life and ability to complete daily activities [7,19]. As an approach, MILRs have been found to result in less associated postoperative pain than resections performed via thoracotomy incisions.

In other surgical specialties, [20] post-discharge opioid consumption was predictive of opioid usage on the day before discharge. The predictive value of opioid usage within 24 hours of discharge has been reiterated in the colorectal, pediatric, and gynecologic surgical literature [21–23]. The role of inpatient use on post-discharge opioid consumption in patients undergoing thoracic surgery procedures has not yet been evaluated.

Analyzing the factors associated with chronic opioid use after a MILR is critical to inform prescription patterns, identify high-risk patients, and understand the clinical variables associated with becoming a chronic opioid user. Therefore, our study aims to identify factors associated with chronic opioid use after MILR, bridging this gap in the literature. We hypothesize that there is a positive association between the usage of opioids in the inpatient setting and the length of post-operative outpatient opioid usage.

Materials and methods

Results

Patient characteristics

A total of 376 patients underwent MILR between March 2019 and May 2022 and met inclusion criteria, with 83.5% of patients having no opioid use beyond 30 days postoperatively, and 16.5% continuing opioid use at 30 days. Among these, 10.1% had opioid use at 60 days, and 8.5% at 90 days. The cohort was 39% male with a mean age of 64.6 years old and 88.8% of patients were white. The mean body mass index (BMI) for the cohort was overweight at 29.1 kg/m². There was no significant difference in gender, age, race, or BMI amongst the opioid use subgroups (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic, preoperative, and operative characteristics for the full cohort and stratified by opioid use duration.

https://doi.org/10.1371/journal.pone.0325354.t001

The most common comorbidities were hypertension, coronary artery disease, diabetes, major psychiatric disorder, past or current smoking status, and atrial fibrillation. The least common were cerebrovascular accident (CVA), pulmonary hypertension, and deep vein thrombosis/pulmonary embolism (DVT/PE) (Table 1). The majority of patients in the cohort had at least one comorbid condition (64.1%) and most (66%) underwent anatomical lung resections.

Intraoperative and postoperative variables

The mean length of anesthesia was similar across the groups, with no significant difference in length of stay (LOS) between those who used opioids beyond 30 days and those who did not (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Postoperative outcomes and opioid use for the full cohort and stratified by opioid use duration.

https://doi.org/10.1371/journal.pone.0325354.t002

Within the first five postoperative days (PODs), opioid consumption was significantly associated with chronic opioid use beyond 30 days. Mean MME consumption on POD one was higher in patients with opioid use at 30, 60, and 90 days, compared to those who discontinued opioid use within 30-days following surgery (Fig 1) (Table 2). This trend continued across the first five PODs, with significant differences in MME consumption over POD two through five (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Mean morphine milligram equivalents (MME) use for each subgroup by postoperative day.

Postoperative complications were not significantly different between the groups of opioid usage. Patients with opioid usage beyond 30 days had a slightly higher rate of complications, including prolonged air leak and readmission, though these were not statistically significant (Table 2). This trend in higher complication rate included any post-operative complication overall, for which the 60-day opioid use group had the highest at 42.1%. No significant difference in unanticipated postoperative invasive procedures or 30-day readmissions were observed, although these complications were in a higher proportion of patients in the groups with opioid usage beyond 30 days (Table 2).

https://doi.org/10.1371/journal.pone.0325354.g001

Chronic opioid use

Multivariable logistic regression analyses identified the MME quantity consumed on the day prior to discharge as a strong predictor of opioid use at 30, 60, and 90 days postoperatively. For each 10-unit increase in MME on the day prior to discharge, the odds of opioid use beyond 30 days increased by 21% (Fig 2) (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Multivariable logistic regression models fit predicting opioid use at 30-, 60-, and 90-days after discharge.

https://doi.org/10.1371/journal.pone.0325354.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Forest plots of logistic regressions predicting opioid use at A) 30 days, B) 60 days and C) 90 days.

https://doi.org/10.1371/journal.pone.0325354.g002

Similar associations were observed at 60 days with a 20% increase and 90 days with an 18% increase (Fig 2) (Table 3). LOS per one day increase was not a significant predictor in any of the models, nor was anatomic resection. Higher MME use in the immediate postoperative period, particularly on the day prior to patient discharge, was a significant predictor of chronic and long-term opioid use in this cohort.

Discussion

In a retrospective analysis of postoperative opioid usage pattern after minimally invasive lung surgeries, we identified a significant association between postoperative in-hospital MME use and chronic (>30 day) opioid use, with a dose-response effect. We found that patients who consumed higher MME in the first five postoperative days, especially on the day prior to hospital discharge, were significantly more likely to continue opioid use at 30, 60, and 90 days postoperatively. The dose-response relationship is consistent with previous studies that have documented the association between perioperative opioid exposure and prolonged opioid use [1,4,10,20,26]. These findings emphasize the need for early and effective postoperative pain management. Importantly, neither LOS nor postoperative complications were significantly associated with prolonged opioid use, highlighting the central role of perioperative opioid consumption as a modifiable risk factor for chronic and long-term use. Recognition of the association is valuable in tailoring postoperative prescription guidelines.

Our findings align with prior literature that patients exposed to higher doses of opioids during the inpatient stay were more likely to continue opioid use beyond the immediate recovery period [20,21,27]. Many previous studies in thoracic surgery focused on pain after thoracotomy incisions [11,12]. Our results suggest that even in minimally invasive surgeries, where PTPS is not a concern, there is a clear association between early postoperative opioid consumption and longer-term use. Our results further highlight the need for careful management of acute postoperative pain, particularly in high-risk patients, such as those with higher immediate postoperative MME use.

Prior literature has identified associations between postoperative opioid use and younger age, male sex, lower socioeconomic status, specific comorbidities (including diabetes, heart failure, and pulmonary disease), as well as preoperative history of drug abuse, alcohol abuse, depression, and benzodiazepine or antidepressant use [4,27]. In our study, MME use was the only significant predictor, with other patient characteristics and operative factors not having a significant degree of association with chronic and long-term opioid use. Anatomic resections approached but did not reach significance (P = 0.07) in the group using opioids at 30-days, however this factor became even less significant at 60-days and 90-days and was not a predictive factor in the multivariable models. Similarly, opioid use was not associated with the length of operative time. The extent of resection and time spent in the operating room were not found to be relevant to the development of chronic pain, which is likely more so related to pain pathways and patient-specific pain perception.

Further evaluation of other potential variables in predicting chronic opioid use is important to assist thoracic surgeons with preoperative planning to optimize pain management for patients. The potential predictive value of these variables has important implications as patients undergoing pulmonary resections frequently present with medical comorbidities, especially pulmonary disease, which has been hypothesized to increase sensitivity to the respiratory depressant effects of opioids, potentially increasing overdose risk [28].

The clinical implications of this study include targeting early opioid use – especially on the day prior to discharge – as a potential strategy to reduce chronic opioid use in MILR patients. Patient-centered, responsible opioid prescription guidelines have been previously evaluated in multispecialty surgical patients, including thoracic patients, based on opioid use the day prior to discharge [29]. Pre-discharge pain management protocols should be optimized to provide education to patients and minimize dependence on opioids without compromising pain control. Some interventions that have been beneficial include multimodal analgesia, preemptive early outpatient intercostal nerve blocks, and cryoablations to decrease chronic opioid usage in high-risk patients [30–32].

Given the ongoing opioid crisis in the United States, where opioid-related deaths make up the majority of overdose mortality, the implications of these findings are substantial [5,6]. Our study suggests that the current perioperative pain management protocols, even those employing multi-modal approaches, may still contribute to the risk of chronic opioid use. Particularly troubling is that a significant proportion of the opioid-naive patients in our study had long-term use of opioids post-surgery. The association between higher in-hospital opioid consumption and chronic use suggests that further refinements in pain management may be necessary.

An important consideration that was not captured in our study is the relationship between patient-reported pain scores and postoperative opioid use. As a retrospective study, we were limited by inconsistent documentation of pain scores after discharge, which prevented their inclusion in the analysis. As a result, the study cannot determine whether higher MME usage reflected greater pain or differences in prescribing practices. Future prospective studies should incorporate standardized pain assessments to distinguish between these possibilities and clarify whether patients with higher opioid requirements are experiencing more significant postoperative pain or are at risk due to prescribing patterns alone.

As a single-center retrospective cohort study, our findings may not be broadly generalizable to other institutions with different perioperative pain management protocols. The majority of our cohort was white, which may limit the applicability to other populations. Additionally, while our multivariable logistic regression helped control for several factors, unmeasured confounders may still influence the observed associations, such as preoperative pain levels and socioeconomic factors not included as variables in the analysis. The absence of consistently recorded pain scores also limited our ability to assess whether higher opioid use was driven by patient pain versus provider prescribing behavior. Finally, reliance on prescription data to define opioid use, while a reasonable surrogate for opioid consumption, may not capture actual patient use of prescription medication [33,34].

Replicating these findings in larger, multi-center, prospective studies and examining the impact of additional perioperative factors on chronic opioid use would be valuable. Future areas of study should focus on assessing the efficacy of preemptive early outpatient interventions, such as intercostal nerve blocks and cryoablations, in reducing the need for opioids after discharge. Additionally, exploring alternative pain management strategies that lower in-hospital opioid administration could be critical in reducing chronic opioid use. The potential of enhanced recovery after surgery (ERAS) protocols in decreasing opioid use should also be investigated, as early data suggests these protocols can lead to better outcomes with lower opioid requirements [35].

Conclusion

Immediate postoperative narcotic usage is strongly associated with chronic opioid dependence after minimally invasive lung resections. The strong association between in-hospital MMEs and continued opioid use at 30, 60, and 90-days postoperatively emphasizes the need for improved pain management protocols. Future research should continue to explore alternative analgesic strategies and their potential to reduce the risk of chronic opioid use, contributing to efforts to mitigate the opioid crisis.